Wolf's Canon Rewrite

Inspiration

Inspired by a classic fairy tale story and modern isekai media, Wolf's Canon Rewrite lets YOU live the story as the villain.

What it does

You must venture through the pixelated woods to reach the cottage, "acquire" your grandmother disguise, and successfully deceive Little Red Riding Hood. Featuring:

LLM-Powered NPCs: Engage in conversations with the Gnome and Little Red through an LLM-integrated dialogue system that reacts to your every lie.
Branching Fate: Your actions determine your ultimate destiny—successfully deceive Little Red to return to Earth, or fail and stay bound to the original narrative.
Pixel Art & Animation: Features a consistent pixelated art style with frame-by-frame animations.
Retro UI/UX: Features typewritten dialogue, a classic blinking cursor, and "Press Enter to Skip" mechanics
Interactive World: Explore the world with responsive keyboard controls, collision detection, and dynamic NPC interactions

How we built it

AI & Logic: Built using in Pygame using the Groq API (Llama-3.3-70b), with asynchronous threading implemented to ensure the game loop never freezes while the AI thinks. Each NPC is operated with a unique system prompt that governs their behavioral logic and status tracking for the interaction.
Artwork: Free assets sourced online or self-composed with Affinity Designer. Canva also was used for further editing and layout construction.

Challenges we ran into

Code Refactoring: Moving from a monolithic main.py to a modular architecture was essential as the game's complexity grew, and this remains as an area for future improvement.
AI Prompting and UI Integration: Mastering prompt engineering to ensure reliable NPC behavior was a major hurdle, as was integrating that real-time AI output into a visually cohesive, typewriter-animated interface.
Loop Synchronization: We used asynchronous threading to prevent Groq API latency from freezing the 60 FPS game loop while NPCs "thought."

Accomplishments that we're proud of

We successfully built a functioning game from start to finish, with a complete experience combining complex game logic, music, animations, and visuals.

What we learned

We gained experience with the Pygame ecosystem, developed proficiency with game loops, event handling, and collaborative version control using Git. We also refined our teamwork, and we learned to use prompt engineering to ensure our AI characters behaved predictably within the game's story and logic.

What's next for Wolf's Canon Rewrite

Scalable Architecture: Further refactoring of the codebase to support maintainability and easier extensibility.
Deepening the Forest: Designing additional NPC stages and encounters to expand the player’s journey and flesh out the world before reaching the cottage.
Branching Narratives: Implementing a wider array of alternative endings to reward player creativity and maximize replayability.
Polishing the Experience: Elevating the intro and loading screens with richer animations and interactive elements to boost player engagement from the very first frame.